Internal Displacement To Urban Areas: The Tufts-idmc Profiling Study (case 2: Abidjan, Cote D'ivoire)

Internal displacement to urban areas: the Tufts-IDMC profiling study

CASE 2: ABIDJAN, CÔTE D’IVOIRE

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study Abidjan, Côte d'Ivoire: Case 2 September 2008 Karen Jacobsen, Feinstein International Center, Tufts University in collaboration with Internal Displacement Monitoring Centre, Geneva

In many conflict zones today, the targeting and uprooting of rural populations and their forced displacement is an integral part of the war strategies of rebel or government forces. Notable recent examples include Sudan, northern Uganda, Colombia, Côte d'Ivoire, Burma and Somalia. Many of these displaced people flee across borders to become refugees, but even more become internally displaced and a large and growing proportion migrate to the urban areas and particularly the capital of their own countries. Unlike internally displaced people (IDPs) in camps who are more easily identified and assisted, IDPs in urban areas comprise a hidden population, and aid agencies and governments have difficulty identifying them and understanding their experience relative to the urban population amongst whom they live. Relatively little is known about their precise numbers, demographics, basic needs and protection problems. Donor governments and humanitarian organizations have recognized this information gap, and in 2006, the Norwegian Refugee Council’s Internal Displacement Monitoring Centre commissioned the Feinstein International Center to conduct a research study that would address this gap. The study had three main objectives: to develop research tools to be used for profiling urban IDPs, including to make population estimates;

•

to generate comparative data on IDPs and non-IDPs in urban areas— including demographic and livelihood characteristics, access to serv-

•

1

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

ices, economic integration, and whether the assistance and protection needs of idps differ from that of non-idps; • to use the data to work with governments and humanitarian organisations to develop programs and advocacy strategies that assist idps and protect their rights. The study took place from 2006-2008, in three urban locations: Khartoum, Sudan, Abidjan, Côte d’Ivoire and Santa Marta, Colombia. Surveys were conducted in each city, and the outcome was a tested profiling tool, a full report, and three case studies. These outputs can be found at www.internal-displacement.org or http://fic.tufts.edu. For information on the studies, please contact the author at Karen. [email protected]. For more information on the idp situations in Sudan, Côte d’Ivoire, and Colombia please visit idmc’s website at www.internal-displacement.org.

2

Abidjan, Côte d’Ivoire: Case 2 of the Tufts-IDMC Profiling Study of IDPs in Three Urban Areas Karen Jacobsen, Feinstein International Center, Tufts University in collaboration with Internal Displacement Monitoring Centre, Geneva

executive summary

A

fter a series of political crises, Côte d’Ivoire’s armed conflict erupted in 2002, leading to the division of the country into government and rebel-controlled areas. Both sides committed grave human rights abuses and hundreds of thousands of civilians fled attacks on villages, kidnappings, and severe persecution based on ethnicity, nationality, or political opinion. Especially targeted were northern nationals and immigrants from Burkina Faso, Mali, and Niger. Many internally displaced people migrated to urban centers such as Abidjan, Grand Bassam, and Yamoussoukro, but in urban centers they were subject to government harassment, including, in Abidjan, demolition of housing and attacks. Three peace agreements brokered between 2003 and 2005 failed, but the signing of the Ouagadougou Peace Agreement in March 2007 raised hopes for a return to stability, and many idps have spontaneously started to return to their areas of origin. Effective monitoring of idps in Abidjan has been hampered by their invisibility. unfpa and unhcr carried out surveys in 2005 and 2007 respectively; however, these did not cover all of Abidjan and were not representative.We made Abidjan a case study to address the need for information about idps and because the city met our study’s criteria. Initial planning for the survey began in March 2007, and the survey was conducted in June 2007. methods In accordance with the overall study methodology, we generated a random, two-stage, systematic sample with a target of one thousand households. For the first stage, using a sampling method known as “probability proportional to size” and the 1998 census list of households, we selected 50 quartiers from the 270 quartiers making up the ten communes of Abidjan. In the second stage, each of the 50 selected quartiers was divided into quadrants, and five

3

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

households were randomly selected across each quadrant for a total of twenty households in each quartier. Our final sample size was 976 households.We created gis maps of the selected quartiers using a combination of Google Earth and commercial maps. These maps assisted the supervisors in navigating around the city, and gis methods were later used to map our survey data. In order to weight the sample appropriately, the study’s methodology requires stratification of each target city into areas of high to low idp density, based on the local knowledge of key informants. However, in Abidjan idps are settled throughout the city and the complexity of the habitat pattern made it difficult to estimate the idp population density of each of the 270 quartiers. As a result we did not stratify the city for the purposes of sampling. Survey data were collected over two weeks in June 2007. Partially cleaned data were sent to Tufts in Boston for cleaning and analysis. We conducted secondary analysis of our data to determine who were idps, and then compared this sub-group with non-idps. survey findings Migration to Abidjan Of our respondents, 35% said they were born in Abidjan, 14% had lived there since childhood, and 30% had arrived in Abidjan since 1995. Our question, “Why did you come to Abidjan?” was answered by 640 (66%) respondents, indicating that 66% of our sample had migrated to Abidjan. The main reasons for coming were finding work (30%), escaping conflict (13%), seeking education (27%), and joining their family (17%). The number and rate of arrivals increased every year from 1948, and the average number of arrivals almost doubled between 2000 and 2006. The spike in arrivals in 2003 and 2004 was probably caused by the influx of people displaced by the conflict that began in 2002. A significantly larger proportion of respondents from the conflict-affected regions of Vallée du Bandama, 18 Montagnes and Moyen Cavally came after 2001. Indicators and Estimates Our survey did not seek out idps. Using secondary analysis of our data, we identified idps based on their reported experience of displacement and when they had come to Abidjan. idps were designated as those who said they had been forced to move or evicted for reasons related IDP

4

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

to government relocation or because of the conflict (“la guerre”), and who had arrived at their current address after 2001 (the conflict broke out in 2002). Of our respondents, 95, or 9.7%, met our criteria for being idps. Our confidence interval is 1.85%, which gives us an expected range of 7.85-11.55. Thus we expect that idps comprise between 7.9% and 11.6% of the urban population. In terms of place of origin, 13.7% of our idps came from Lagunes; 48.4% came from the three conflict regions of Moyen Cavally, Bandama and 18 Montgnes; and 37.0% came from the fifteen other regions of Côte d’Ivoire. idps were found in all the communes of Abidjan except Treichville. The communes with the highest proportion were Youpougon, and Abobo, and those with the lowest proportion were Port Bouët and Marcory. The main ethnic groups of the idps were the Baoulé (26%), the Guéré (20%), and the Yacouba (14%). Making Population Estimates of IDPs Updated estimates from the last census in 1998 put the population of Abidjan at approximately 3.8 million in 2006. Using our 9.7% proportion of idps, we estimate that idps in Abidjan number in the range of 299,937–440,414. (Using a confidence interval of ±1.85%, and our expected range of 7.9 and 11.6%) This compares with the unfpa/ensea study, conducted in 2005, which estimated a total number of 495,783 idps for Abidjan. Our study occurred two years later, and in the context of return movements that had already started. Once the figures for the next census are available, a better estimate can be made. Our estimate is likely to be conservative because idps are also hosted in non-idp households, but we did not include them in our estimate because unfortunately, we realized the importance of this factor only after we had begun the survey, and thus could only ask the question of a subset of respondents. Comparing IDPs and Non-IDPs in Abidjan We compared idps and non-idps regarding their housing and education, their movement and experience with forced evictions, their employment, difficulties they experienced, and intentions regarding future movements, including returning home. Statistically significant findings are as follows:

5

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

• Demographics and Household 54% of all our respondents were men.The average age for men was 37, and for women 35. Half of the sample (49%) was married or cohabiting with a partner. Households had an average of two children, and 17% of households did not have children.The average (mean) number of household members was 8.2. The total number of household members represented by our 976 respondents was 7,953, of whom 2,781 were under 18 years of age. idps did not differ statistically from non-idps in any demographic respects (age, gender, marital status), but idps tended to have a slightly higher number of children in their households. • Housing quality Most of our respondents (69%) lived in brick houses, with another 21% in concrete houses. Seven percent said they lived in temporary dwellings (or shacks), and three percent lived in dwellings made of mud. idps and non-idps did not differ statistically in their type of housing. A slightly larger proportion of idps (9.5%) lived in temporary shelters than did non-idps (6.7%). idps were more likely to be renters than owners, and more likely to be hosted by others (usually family). • Education Almost 12% of the total sample was illiterate, four percent had religious education, 19% had completed primary school, 35% had completed secondary school, and 28% had some university education. In exploring idp vs. non-idp education levels, we first split our samples by gender. idp men and women tended to have slightly more education than non-idps. Illiteracy rates were higher for non-idp men and women, and idp men had more university education than non-idp men (44% vs. 35%). Secondary school levels were about the same, around 35%, for all groups. Larger proportions of both idp and non-idp women had only primary school compared with men. Women in both groups were significantly less likely to have had university education: only 19% of non-idp women and 12% of idp women had any university, compared with 35% (non-idp) and 44% (idp) of men. • Employment Of our sample, 16% said they were in full-time employment, and 16% were unemployed. 11% were in part-time or casual employment, and 8% were housewives. 15% said they were students. idps were more likely than non-idps to be both unemployed and students, while nonidps were more likely to be “housewives,” but the two groups were

6

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

similar in other employment categories. Some 17% reported full-time employment, and 17% said they were self-employed. As with education, employment patterns were differentiated more by gender than by idp status. Both idp and non-idp women were more likely to be self-employed than men, and much less likely to have full-time or parttime/casual employment or to be students. • Difficulties experienced in Abidjan We found little difference in standard of living indicators, including access to potable water, distance from facilities, respondents’ expressed difficulties, and whether respondents had received assistance from the government or aid organizations in Abidjan. Where there were differences, idps generally appeared to be worse off. The similarity of living conditions is probably explained by the fact that idps are living within host communities, sometimes even in the same houses, and while they may be slightly poorer, probably not enough to show up statistically. Access to potable water: idps were slightly more likely to use water vendors and less likely to have a house connection, but this was not a statistically significant difference. Access to social services: Most people (93% of the total sample) lived within a kilometer both of their children’s school and of some means of transportation. idps tended to live somewhat further away from a health facility and from a police station. Household difficulties: Almost all respondents mentioned the difficulties of finding work, and idps were more likely than non-idps to feel that their living area was unsafe (52% versus 42%). Many respondents experienced problems with their water supply and said the drainage ditches created sanitation problems. Other problems were distance from transportation (more than 90%), and general sanitation of communes. Mobility within Abidjan: idps were significantly more likely to have re-located their residence within Abidjan. 57% of idps had lived in other parts of Abidjan previously, versus 36% of non-idps. This could be a reflection of the demolition campaign and other targeting of idps by the government during 2002.

7

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Assistance from government or aid agency: Relatively few respondents, just 21 in total (2.2%), had ever received assistance from the government or aid agencies. idps were significantly more likely to have received assistance however. Of the 21 recipients, nine were idps. • Identity Documents The possession of identity documents by idps is a key protection concern. Lack of documents makes it difficult to establish one’s right to be in a place, to claim services, or to lay claims to land. In Côte d’Ivoire, birth certificates are important for accessing one’s rights as a citizen. During flight and displacement, personal documents are often lost, and idps can have difficulty replacing them. We found that idps were less likely to have a birth certificate (84% vs. 94%), but other documents—nationality certificates, driver’s licenses and passports—were held at similar rates. idps held national identity cards and voter cards at higher rates than non-idps. • Future Migration Intentions idps were significantly more likely to want to return home than nonidps (44% vs. 4%), and less likely to want to remain where they were in Abidjan. Whereas 77% of non-idps wanted either to remain where they were or move elsewhere in Abidjan, this was the case for only 34% of idps. Of all our respondents only two (less than 1%) said they wanted to migrate to North America or Europe. In sum, while idps and non-idps experience the same stresses related to urban poverty and lack of adequate infrastructure, and share similar demographic characteristics and household situations, there were some notable differences: in their employment and housing situations, in the kinds of identity documents they possess, and in their desire to return home or to leave Abidjan. These differences suggest that idps are more vulnerable—they experience more structural insecurity and find it harder to lay claim to their rights as citizens. Key areas of vulnerability include identity documents, housing, and employment. It is worth exploring further to see whether idps experience discrimination in the areas of housing and employment. Some aid agencies in cooperation with the government have started helping idps obtain or replace birth certificates and other identity documents.

8

table of contents acknowledgements

11

displacement context: conflict and internal displacement in côte d’ivoire 14 the tufts-idmc study methodology and challenges

19

1. demographic and household characteristics of all respondents 2. migration to abidjan and the impact of the conflict 3. idp indicators 4. estimates of idp population in abidjan 5. comparison of idps with non-idps in abidjan 6. analysis and conclusion

23 25 29 33 34 41

appendix a: abidjan communes and quartiers with no. of survey respondents

43

appendix b: ethnicity of idps and non-idps in abidjan

44

9

index of tables, maps, and charts map of côte d’ivoire internal displacement in côte d’ivoire, internal displacement monitoring centre, April 2008 map of surveyed area of abidjan table 1.1 demographic characteristics chart 1 main ethnic groups chart 2.1 reasons for coming to abidjan chart 2.2 arrivals who came to escape conflict, by year chart 2.3 place of birth and residence before coming to abidjan table 2.1 place of birth and when came table 3.1 idps by place of birth table 3.2 idps by COMMUNE in Abidjan chart 3.1 distribution of idps in COMMUNES of abidjan table 3.3 ethnicity of idps chart 3.2 ethnicity of idps in abidjan table 4.1 non-idps hosting idps by COMMUNE table 5.1 demographic comparison of idps and non-idps table 5.2 education of idps and non-idps table 5.4 employment in abidjan: idps vs. non-idps chart 5.4 employment in abidjan: idps vs. non-idps chart 5.4a idp and non-idp employment by gender chart 5.5 access to potable water table 5.6 possession of identity documents table 5.7 intentions about future mobility/migration chart 5.7 intentions about future mobility/migration

10

12 13 22 24 25 26 26 27 28 30 31 31 32 32 34 35 35 34 34 37 38 39 40 40

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

acknowledgements We extend our thanks and appreciation to the team of enumerators, supervisors and data entry people in Abidjan: • Team Leader: Eric Levron • Field Supervisors: Alice Katia, Guy Charles Dibahi, Jean Joseph N’gotta • Enumerators: Adelphe Yoa Dimitri, Henriette N’guessan, Anselme Sohou, Amara Bamba, Carole Diomande, Pacome Danzo, Jean Martial Koffi, Alimata Kone N’gadio, Anselme Kadja, Fabienne Konan, Anita Liade, Omer Edia, Roland Abole, Eric Bladi, Sylvain Kouassi, Amy Kone • Data Entry: Aimée Ossoro Ahoussiyo, Fatoumata Toure • Additional Assistance: Tano Kouassi Honoré (Driver), Eric Demers (Country Director, nrc), Lin Anett Haugsland (Finance and Administration Manager, nrc) We received extensive assistance from the research team at idmc, particularly Marzia Montemurro and Arild Birkenes. For gis, GPS, and maps, we thank Patrick Florance (Academic Technology, UIT, Tufts University).

11

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Map of Côte d’Ivoire

12

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Internal displacement in Côte d’Ivoire, Internal Displacement Monitoring Centre, April 2008

13

displacement context: conflict and internal displacement in côte d’ivoire1

T

he conflict that broke out in 2002 took place after a history of relative peace, prosperity, and stability. After independence from France in 1960, Côte d’Ivoire enjoyed political stability and economic prosperity, as a result of the political pragmatism of President Félix Houphouët-Boigny’s autocratic regime and a favorable economic environment. Reliance on foreign labor was key to the country’s economic development, as was the introduction of private landownership. A negative economic trend began at the end of the 1980s, set off by a sharp drop in the commodity prices of Côte d’Ivoire’s main exports, coffee and cocoa. As standards of living began to decline, tensions between locals, non-national West African workers, and internal economic migrants increased. Opportunities arose for ethnic hate politics. In 1995, in a bid to hang on to power, Houphouet-Boigny’s successor, Henri Konan Bédié introduced the concept of “Ivoirité,” a virulent form of nationalism, in order to deny Ivorian citizenship to his main political rival, Alassane Ouattara and provide a pretext to bar him from participating in elections. “Ivoirité” was used both to exclude political opponents and to mobilise popular support against people of foreign origins. About one quarter of the Côte d’Ivoire population of 17 million was classified as foreign, although many had been in the country for two or more generations and by law were entitled to Ivorian nationality. Residence rights of millions of labor migrants were undermined, especially in the north and west of the country, and they became increasingly disenfranchised as economic crisis deepened in the late 1980s and 1990s. A protracted political crisis began on December 24, 1999, when the army, led by General Robert Gueï, overthrew the government of Konan Bédié in the country’s first coup. Building on the theme of “Ivoirité,” Gueï introduced stricter eligibility requirements for the presidential election of October 2000, which was won by Laurent Gbagbo. Like his predecessors, Gbagbo made the issue of nationality central to his political agenda and failed to resolve the growing ethnic and religious divisions across the country. Armed conflict broke out in September 2002 when three rebel groups, who later formed a 1 This section is primarily drawn from two idmc reports: More IDPs return as the peace process moves forward, idmc, 10 April 2008 and Internal displacement in Côte d’Ivoire: a protection crisis, Global idp Project, 7 November 2005. See also Siendou A. Konate, “The Politics of Identity and Violence in Côte d’Ivoire,” West Africa Review: Issue 5, 2004. http://www.africaresource.com/war/issue5/konate.htm

14

military-political alliance called the Forces Nouvelles, attempted a failed coup. Intense fighting left the rebels of the Mouvement Patriotique pour la Côte d’Ivoire (MPCI) in control of much of the center and the predominantly Muslim north of the country, with government forces holding the largely Christian and more prosperous south. Hundreds of thousands of civilians were displaced. At least 200,000 people were estimated to have fled the rebel-held northern town of Bouaké and many sought refuge in the urban centers of Abidjan, Grand Bassam, Yamoussoukro, Duékoué, Toulepleu, and Guiglo. Further displacements occurred later that year when two new rebel factions emerged in western Côte d’Ivoire. The center, north, and west of the country was affected by armed conflict, the south and west were made insecure by the presence of armed elements, and the south-west and western areas were characterized by land or inter-communal conflicts. Many conflict-displaced people eventually found their way to Abidjan, mixing with other migrants. Violence persisted in the country with civilians fleeing attacks on villages, systematic human rights abuses, kidnappings, and severe persecution based on anti-foreigner sentiment.2 Especially targeted were northerners and immigrants and residents from Burkina Faso, Mali, and Niger. By 2005 most idps were living in overburdened host communities in the government-controlled south of the country. Less than 10,000 idps lived in established camps or centers.3 During this period of violent upheaval, thousands of idps came to Abidjan. In 2002, the government set about demolishing homes in the city’s shantytowns in an effort to weed out insurgents, rendering thousands homeless. The main targets of the demolition were non-nationals whom the authorities accused of supporting the rebellion. In March 2004, government troops and their allied militia killed at least 120 people during an opposition march in Abidjan. Grave human rights abuses, including killings, disappearances, torture, and destruction of property have been committed against civilians in both the government-held south (particularly by pro-government youth militia) and in the rebel-held north, as well as in the Zone of

2 According to article 6 of the Code de la Nationalité (Law n. 61 - December 14, 1961, modified by the law n 72-852 of December 21, 1972), a person who has either an Ivorian mother or father is entitled to Ivorian nationality. The naturalization or restoration of one’s nationality can also be granted by decree upon request and after an inquiry. 3 Côte d’Ivoire’s only purpose-built idp camp, in the western town of Guiglo, housed some 7,900 idps, mostly Burkinabé immigrants, in two sites. Although intended as a temporary transit center, it was officially closed only in August 2008.

15

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Confidence—the zone between government—and rebel-held areas where UN and French peacekeepers established 17 military observation posts. Crimes have been committed with impunity on all sides, reinforced in part by the UN’s failure to impose sanctions against individuals as provided for under the November 2004 Security Council Resolution 1572,. International efforts to end the civil war had limited success.Three peace agreements brokered between 2003 and 2005 failed, and violence persisted in the country. The political impasse became more intractable when it became clear that presidential elections scheduled for October 30th, 2005, would not go ahead. The planned disarmament of both rebel forces and pro-government militia was shelved as the parties continued to wrangle over legislative reforms relating to citizenship and land tenure. In addition, opposition leaders rejected a proposal for South African mediation, alleging they would be biased towards the government.The Security Council finally supported African Union proposals to allow Gbagbo to remain in power for up to twelve months beyond the end of his mandate, as long as he delegated various powers to a new and more powerful prime minister—Charles Konan Banny, governor of the Central Bank of West African States. In December 2005 Banny formed a transitional government that was faced with the formidable task of organizing disarmament, identification, and elections within just ten months. The fact that the long-awaited presidential elections could not be held by the October deadline was finally accepted by the UN Security Council, and in a November 2006 resolution the Security Council formally extended Gbagbo’s mandate on condition that the office of prime minister be strengthened. The signing in March 2007 of the Ouagadougou Accord by President Gbagbo and Guillaume Soro, Secretary-General of the Forces Nouvelles, bolstered hopes. Named after the capital of neighbouring Burkina Faso, where it was signed, the agreement calls for the abolition of the zone of confidence, the removal of the observation posts, and the formation of a new transitional government involving the two sides, the merging of Government and Forces Nouvelles troops, and the holding of free and fair national elections. An assassination attempt in June 2007 on Prime Minister Soro, exrebel leader of the Forces Nouvelles, underlined the fragility of the peace process, but there has been some progress. Presidential elections have been scheduled for November 30th, 2008. By the end of July 2008, the UN had dismantled all military posts. Its Office for the Coordination of Humanitarian Affairs noted the progress towards reconciliation but warns that “the disarmament, demobilisation and rehabilitation pro-

16

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

cess has not moved forward as anticipated, nor has the creation of an integrated army. There have been the beginnings of the redeployment of local state authorities (notably Prefects and Sub-Prefects) back to the zones formerly controlled by the rebels, but the redeployment of Government officials has been slow and social services remain heavily overstretched.”4Attacks on civilians by armed men are increasingly common, and the government has failed to bring large parts of the country under its control. Former rebels continue to tax trade and transport in the north.5 The International Crisis Group warns that: The possibility of a return to violence is not to be overlooked as effective disarmament of armed forces and militia, the successful identification of voters, and the redeployment of the state administration to the North of the country remain tough challenges to overcome before credible elections can take place and the country can be reunified. The combination of a population whose economic conditions have been deteriorating steadily and a minority from both conflicting parties, enriching themselves over the no peace-no war situation might be an additional threat to Côte d’Ivoire’s stability.6

In the volatile western and northern regions, idps and other vulnerable groups continue to have little access to basic social services, particularly health care, water/sanitation, and education. Malnutrition rates remain high especially among children under five and in landlocked areas in where conflict-induced displacement had impacted on the already weakened crop production systems. In the rebel-held areas, many schools had not been functioning since the outbreak of the crisis in 2002, not least because large numbers of teachers (and other civil servants) were displaced to major towns in the south. The signing of the Ouagadougou Peace Agreement and the gradual redeployment of public servants laid the foundations for gradual improvement in areas of return, and as a mark of positive development, on 17 September 2007, the school year started on the same day nationwide for the first time since the civil war erupted. In the west, long-standing tensions over land between indigenous communities and settlers from neighboring countries including Burkina Faso, Mali, and Niger (as well as Ivorians from northern ethnic groups) remain particularly acute. Inter-community attacks and low 4 UNOCHA, Consolidated Appeal for Côte d’Ivoire 2008 http://ochaonline2. un.org/Default.aspx?alias=ochaonline2.un.org/Côtedivoire 5 UNOCHA, IRIN, March 2008 http://www.irinnews.org/Report.aspx?ReportId=77409 6

http://www.crisisgroup.org/home/index.cfm?id=1235

17

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

level displacement continue unabated. In the north, Forces Nouvelles members still occupy houses owned by idps. Identified by many as a pre-condition for return, repossession of occupied property is still a delicate matter, which is further aggravated by internal differences within the Forces Nouvelles and the lack of an operational judicial system. In general, access to justice remains difficult, especially for the most vulnerable groups including internally displaced people, as access to courts is limited by distance, cost and widespread ignorance of the necessary administrative procedures. Extortion and racketeering are still rampant throughout the country, while sexual and gender-based violence is of major concern (including the exploitation of internally displaced girls by peacekeeping troops).

18

the tufts-idmc study methodology and challenges Today, as at the time of research in early 2007, the main objectives central to sustainable resolution of conflict in Côte d’Ivoire are: • guaranteeing Ivorian nationals the right to exercise their nationality; • ensuring that so-called “foreigners” can exercise another nationality or acquire Ivorian nationality on the basis of their long-term residence in Côte d’Ivoire; • collecting and disseminating accurate data on idps in Abidjan and throughout the country in order to bring about the first objective and inform programming and advocacy. Effective monitoring of idps has been hampered by their invisibility and the complex patterns of their displacement. unfpa carried out a survey in 2005. In February-March 2007, unhcr carried out a profiling exercise in four communes of Abidjan believed to host the highest density of idps. However, that exercise did not cover all of Abidjan and was not representative. nrc and idmc decided to make Abidjan one of the case studies for the urban displacement project to address the need for data, and also because the city offered a good opportunity to meet the objectives of the overall idmc study. Initial planning for the idmc survey began in March 2007, and the survey was conducted during two weeks in June 2007. Owing to time and resource constraints, we were not able to utilize a media campaign on radio, television and in the press to encourage heads of households to participate in the survey. first phase Our consultant initially spent three weeks in Abidjan, working with the Country Offices of unhcr and nrc and the National Institute of Statistics (Institut National de la Statistique - ins) of Côte d’Ivoire, to introduce the study and ensure institutional participation and cooperation. We encountered several challenges. Inadequate coordination with unhcr, who were conducting their idp profiling exercise in

19

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

four communes (using a different methodology)7, led to a four-month delay before our survey could begin. Cooperation with unhcr improved but the delay led to time constraints during implementation of the survey. With unhcr cooperation, we added several questions to our questionnaire reflecting specific issues confronting idps in Abidjan. The lesson is that enough time should be allocated to involve all stakeholders before dates for implementation are fixed. We also attempted to gain access to the Geographic Information System (gis) census data for Abidjan, but this proved difficult because of the malfunctioning of the ins and the high price of obtaining their data. Consequently, we could only use what was publicly available, but the 1998 census data and commercial maps of Abidjan enabled us to conduct the sampling for Phase 2. In order to weight the sample appropriately, the study’s methodology sought to stratify each target city into areas of high to low idp density, based on the local knowledge of key informants. In Abidjan stratification was difficult because idps are settled in all parts of the city. The biggest communes, such as Abobo and Youpougon, have large property rental markets and host many idps. However, the city is not well zoned and residential, administrative, and industrial areas are mixed together. Former Abidjan villages of the pre-urbanisation period as well as shantytowns are found within more upscale residential areas. The complexity of the habitat pattern made it difficult to estimate the idp population density of each of the 270 quartiers. As a result we did not stratify the city for the purposes of sampling, but used a uniform approach of targeting fifty quartiers across the city’s ten communes. During this field phase we also identified a local community-based organization that had worked with idps, who would provide us with assistance and enumerators for the survey. second phase In the second field phase, we conducted the household survey, after adapting and translating the questionnaire for Abidjan, with the help of the Country Offices of unhcr and nrc. In accordance with the overall study methodology, we generated a random, two-stage, systematic sample with a target of one thousand households. For the first sampling stage, using a sampling method known as “probability proportional to size” (PPS) and the 1998 census list of households, we selected 50 quartiers from the 270 quartiers making up the ten communes. In the second stage, each of the 50 selected 7 We compare some of their findings with ours in subsequent sections of this report.

20

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

quartiers was divided into quadrants, and twenty households were randomly selected across each quartier. (For details of the sampling strategy, see Methods Annex.) Working with three teams of five enumerators and one supervisor, interviews were conducted in 3-4 sites per day, averaging 75 interviews per day. We created gis maps of the selected quartiers using a combination of Google Earth and commercial maps. These maps assisted the supervisors in navigating around the city, and gis methods were later used to map our survey data. Data entry (using Excel) was conducted each evening during the course of the survey. A focus group with the enumerators helped assess the validity of the data and the data collection experience and identify issues and problems. In the final phase, the database was sent to the senior researcher in Boston where the data were cleaned, analyzed, and written up. Secondary analysis of data enabled us to determine who were idps, based on our definitional criteria. The data were analyzed using SPSS and Stata. The draft findings and analysis were sent to nrc Côte d’Ivoire, unhcr, and idmc for comment. During April and May 2008, idmc and the senior researcher discussed our findings in workshops held in Yaoundé and Geneva with donors and aid agencies, including unhcr. limitations of the survey data The survey confronted several institutional and methodological obstacles, imposing some limitations on the validity of our data. These limitations are listed below in order to put the findings in perspective. • Our research design utilized a systematic random sample across the whole city, which meant we probably missed “pockets” of idps where they live clustered in certain areas.This means our population estimate is likely to be conservative. • The survey findings and conclusions are based on self-reported data from respondent households. Response bias is a commonly reported limitation of household surveys, and in our case, it is likely that respondents were reluctant to be specific about politically sensitive issues including their ethnicity, their reasons for displacement, the nature of their identification documents, and their place of birth.

21

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

• One of our questions—the number of idps living in the households of non-idps—was only added a week after our survey began, when we discovered the importance of asking this question.Thus it was asked of only 332 respondents, or just over a third (34%). • Some enumerators were new to the task of interviewing and the use a complex questionnaire that required asking about politically controversial issues such as identification documents and place of birth. Enumerators also faced cultural and language barriers with respondent households, as well as some reluctance to be interviewed in certain areas. Despite training of enumerators, these issues may have affected some survey responses. Map of Surveyed Area of Abidjan

22

survey findings

W

e interviewed a total of 976 respondents in 52 quartiers across nine communes of Abidjan. (Due to the low residential population of Plateau we dropped this commune from our analysis.) Our sampling resulted in the distribution of households found in the Map of the Survey Area of Abidjan on p. 22. Our survey response rate was about 85%, and 4% of questionnaires (43 out of 976) were incomplete. Willingness to respond to requests for an interview was lowest in the richer residential areas (such as Cocody). In the poorer areas, most people were willing and even pleased to be interviewed. Main reasons for refusal and non-completion were concerns about safety, lack of time (especially for the women), and boredom with being interviewed (survey fatigue). In Djibi (Abobo), the head of the village (Ebrié community) told our enumerators to stop and quit the village immediately.

1. demographic and household characteristics of all respondents As shown in Table 1.1, 54% of the sample was men.The average age for male respondents was 37, and for women 35. Half of the sample (49%) was married or cohabiting with a partner. Households had an average of two children, with a range of 1-17, and 17% of households did not have children. The average (mean) number of household members was 8.16 (SD 5.25). The modal household size was four members, with a median of seven and a range of 1–46. The total number of household members for all our 976 respondents was 7,953. See Figure 1 for detailed breakdown of household size. About 12% of the sample was illiterate, with another 4% having religious education (church or Koranic), 19% had completed primary school and 35% had completed secondary school. Some 28% had some university education.

23

Table 1.1 Demographic Characteristics (N=976)

Percent male Mean age by gender

Marital status Household size

Children in Household

54 Male 95% CI Female 95% CI % married or living with someone Mean Mode Median Range Total number in households Mean Mode Median No. households with no children Total number children in households

Education

Type of dwelling

Type of employment

Range Illiterate Koranic/church Primary school Secondary school University Concrete Red bricks Mud Temporary Unemployed Part-time + casual Full-time Housewife Student

37 35-38 35 34-36 49% 8.16 (SD 5.25) 4 7 1-46 7953 2.86 (SD: 2.64) 2 2 166 (17%) 2781 0-17 114 (12%) 34 (4%) 183 (19%) 338 (35%) 272 (28%) 201 (21%) 674 (69%) 32 (3.3%) 68 (7%) 157 (16%) 102 (11%) 160 (16%) 75 (8%) 145 (15%)

hosting idps We asked a subset of our sample (332 respondents in five communes) whether they were hosting people who had been forcibly displaced. Of this subset, 11.5% said they hosted idps. Most housed one or two idps, but some housed as many as ten. The average number was three (standard deviation 2.60), for a total of 112 idps housed by 37 households. We can assume that idps are being housed across Abidjan.

24

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

ethnic groups As shown in Chart 1, Côte d’Ivoire’s ethnic groups were fully represented in the sample. The largest ethnic group is the Baoulé (16%), followed by the Bété (9%), and the Agni (7%). Chart 1: Main Ethnic Groups Main Ethnic Groups 18% 16% 14% 12% 10% Series1 8% 6% 4% 2%

al

M

ou Ba

M

le os si in ke Pe Se ulh no uf G o ue re Be te Ag n H i a Ya usa co ub a M or Ab e r Ko on ya k G a Ad ou jo ro uk ro D u io ul a D id a At t Ap ie po lo Eb ri Ab e Ko ba ul y a Ta ngo gb Ba an m a ba ra Lo b O i th er

0%

2. migration to abidjan and the impact of the conflict Abidjan is a migrant city, and there is a constant flow of arrivals and departures. Of our respondents, 35% said they were born in Abidjan, 14% had lived there since a child, and 30% had arrived in Abidjan since 1995. reason for coming As with most migration, people come to Abidjan for four main reasons: to find work or pursue a livelihood, for education, to join their families, or to escape conflict and persecution. Our question, “Why did you come to Abidjan?” was answered by 640 (66%) of our 976 respondents; (the other 326 (34%) were born or grew up in Abidjan). As shown in Chart 2, reasons for coming were divided between finding work (30%), escaping conflict (13%), seeking education (27%), and joining their family (17%). Other reasons given included because their livelihood failed (3%). A small number gave more than one reason, but most gave only one main reason. (These percentages reflect the number of times a reason was mentioned.)

25

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Chart 2.1 Reasons for Coming to Abidjan 35%

30%

25%

20%

15%

10%

5%

0% Find work

Escape conflict

education

join family

livelihood failed

other

year of arrival in abidjan The number and rate of arrivals increased every year from 1948, and the average number of arrivals almost doubled between 2000 and 2006. The spike in arrivals in 2003 and 2004 was probably caused by the influx of people displaced by the conflict that began in 2002. Chart 2.2 shows the flow of all arrivals between 1996 and 2006, and the proportion of those who gave “escape from conflict” as their reason for coming to Abidjan. Respondents who arrived after 2001 were more likely to say that their reason for coming was related to conflict. Of those respondents who said they came for reasons related to conflict, 75% (61 out of 82) arrived in those two years. Chart 2.2 Arrivals who came to escape conflict, by year 60

50

40

No of arrivals Escape conflict

30

20

10

0 1996

1997

1999

2000

2001

2002

2003

2004

26

2005

2006

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

place of birth and where living before coming to abidjan Place of birth. Just over a third (36%) of our respondents were born in Abidjan. As shown in Chart 2.3, migrants were born throughout Côte d’Ivoire’s nineteen regions, including elsewhere in the region of Lagunes (where Abidjan is located), and other countries. The smallest number came from Bafing (<1%) and the largest number from the Vallée du Bandama region (6%). About 8% (74 respondents) said they were born in another country. Living Elsewhere. A proportion of our respondents had been living elsewhere in Côte d’Ivoire other than their place of birth before Abidjan. Of these, those who had been born in Abidjan were least likely to have lived elsewhere (6%), but other regions were distributed more or less similarly to place of birth, except that a slightly larger proportion had been living in Vallée du Bandama (8%), one of the main conflictaffected regions. Chart 2.3 Place of birth and residence before coming to Abidjan 9% 8% 7% 6% 5%

Place birth Livig before Abdj

4% 3% 2% 1%

El

se w

he re

La Su gun e d Co s m oé H l ’ au Ag tSa neb y ss an la d Va Sa ra llé va e ne B s 18 an da M on ma ta gn es La c Su Za s n d Ba zan nd am Fr om a N ’z age iCo r M mo ar é ah ou Ca e va An ly ot O he th er r s co un tr y

0%

Chart 2.3 includes all regions where more than two percent of respondents were born. “Others” in the chart includes the five smaller regions where less than two percent were born, as follows:

27

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Region Moyen-Comoé Bas-Sassandra Worodougou Denguélé Baffing Total

Place of birth 2% 2% 1% 1% 1% 7%

Livig before Abdj 2% 3% 1% 1% 1% 8%

The regions mainly affected by the conflict in Côte d’Ivoire were:8 • Lagunes (where Abidjan is located) • Vallée du Bandama (in the north-central region) • 18 Montagnes (in the west) • Moyen Cavaly. As shown by the shaded rows in Table 2.1, a significantly larger proportion of respondents from the conflict-affected regions came after 2001, and respondents who said they came for conflict reasons were disproportionately likely to be from conflict-affected regions. Thus while only 6% of our respondents came from Vallée du Bandama, they constituted 20% of those who came for conflict reasons and who came after 2001. Just 4% of respondents were from 18 Montagnes but they constituted 17% of those who came for conflict reasons and who came after 2001, and 3% of respondents came from Moyen Cavally but constituted 16% of those who came for conflict reasons and after 2001. Respondents from Lagunes comprised 40% of the total sample, and would not have been part of those who came to Abidjan after 2001 since they already lived there. However 6% of those born in Abidjan were living elsewhere and came to Abidjan after 2001. Table 2.1 Place of birth and when came Place of birth

N and % total sample

Region des Lagunes Region Du Sud Comoe Region de l’Agneby Region du Haut-Sassan Region des Savanes

387 35 43 28 24

40% 4% 4% 3% 2%

% who came after 2001 6% 24 20% 7 12% 5 21% 6 33% 8

8 According to unhcr’s profiling exercise, the main provinces of origins for idps are Vallée de Bandama (40,82%), 18 Montagnes (30,8%), Lagunes (9,90%), and Moyen Cavally (4,27%).

28

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Region de la Vallee de Bandama Region du Moyen-Comoe Region des Montagnes Region des Lacs Region du Zanzan Region du Worodougou Region du Denguele Region du Sud Bandama Region du Fromager Region du N’zi-Comoe Region de la Marahoue Region du Cavaly Region du Baffing Another country Total Total idp areas*

61 16 43 23 28 30 13 25 48 38 22 29 7 74 974 196

6% 2% 4% 2% 3% 3% 1% 3% 5% 4% 2% 3% 1% 8% 100% 20%

33% 20 13% 2 40% 17 26% 6 14% 4 23% 7 23% 3 32% 8 27% 13 21% 8 32% 7 55% 16 0% 0 5% 4

3. idp indicators Our survey did not seek out idps. Through secondary analysis of our data, we identified idps based on their reported experience of displacement and when they had come to Abidjan. idps were designated as those who met the following criteria: • They said they had been forced to move or evicted for reasons related to government relocation or because of the conflict (“la guerre”), • They who had arrived at their current address after 2001 (the conflict broke out in 2002). We did not ask our respondents whether they considered themselves to be idps, but we asked our enumerators to indicate after the interview if they thought the respondent was an idp. Our enumerators were members of the community-based organization that assisted us, and had extensive experience working with the urban population. Their views served as an important validity check. For the most part, their indications matched our analysis. Of our respondents, 95, or 9.7%, met our criteria for being idps. Our confidence interval is 1.85%, which gives us an expected range of 7.85-11.55. Thus we expect that idps comprise between 7.9 and 11.6% of the urban population.

29

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

3.1 origin, distribution and ethnicity of idps Table 3.1 shows that 13.7% of our idps came from Lagunes (both Abidjan and non-Abidjan); 48.4% came from the three conflict regions of Moyen Cavally,Vallée du Bandama and 18 Montgnes; and 37.0% came from the sixteen other regions of Côte d’Ivoire. Table 3.1 IDPs by place of birth Non-idps Place of Birth

Number

Lagunes (Abidjan)

323

idps Number and pct from region 10 3.1%

Lagunes (non-Abidjan)

51

3 5.9%

3.2%

Zanzan

28

0

0

Bas-Sassandra

18

0

0

Worodougou

10

2 20.0%

2.1%

Denguélé

11

2 18.2%

2.1%

Sud Bandama

23

2 8.7%

2.1%

Fromager

44

4 9.1%

4.2%

N’zi-Comoé

33

5 15.2%

5.3%

Marahoue

20

2 10.0%

2.1%

Moyen Cavally

15

14 93.3%

14.7%

Bafing

7

0

0

Sud Comoé

32

3 9.4%

3.2%

l’Agneby

39

4 10.3%

4.2%

Haut-Sassandra

26

2 7.7%

2.1%

Savanes

19

5 26.3%

5.3%

Vallée du Bandama

44

17 38.6%

17.9%

Moyen-Comoé

15

1 6.7%

1.1%

18 Montagnes

28

15 53.6%

15.8%

Lacs

20

3 15.0%

3.2%

Another country

74

-

Total

881

95 10.8%

30

% of idps (n=95) 10.5%

100%

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

As shown in Table 3.2 and Chart 3.1, we found idps in all the communes of Abidjan except Treichville. The communes with the highest proportion were Youpougon, and Abobo, and those with the lowest proportion were Port Bouët and Marcory.9 Table 3.2 IDPs by commune in Abidjan Attcbe CoMarKoucody cory massi Num6 12 3 7 ber of idps 7.5 9.6% 3.8 5.7% % idps in commne % of all 6.3 12.6% 3.2 7.4 idps

Adjame

Abobo

Yopogn

Trchvile

10

24

30

12.7

12.5

16.4

0

3.8

10.5

25.3

31.6

0

3.2

3

Chart 3.1 Distribution of IDPs in communes of Abidjan 35

30

25

20

15

10

5

0 Attecoube

Cocody

Marcory

Koumassi

Adjame

Abobo

Yopougon

Treichville Port-Bouet

3.2 ethnicity of idps As shown in Table 3.3, ethnic groups with the highest proportion of idps are the Baoulé (26% of our idps were Baoulé, and of those who identified themselves as Baoulé, 17% were idps), the Guéré (20% of the idp group and 38% of Guéré respondents) Yacouba (14% of the 9 Our findings differ from those of unhcr’s profiling study (2007), which found: “Among the idp households who took part in the survey, 62% were in the commune of Yopougon, 26.3% in Abobo et 7.7% in Port-Bouët. The communes of Treichville and the town of Grand Bassam had respectively 1% and 3% of the total interviewees.”

31

PrtBët

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

idp group and 43% of Yacouba respondents). By contrast, the Agni comprised 6% of the idp sample (which was close to their proportion of the total sample), and only 9% of the Agni “qualified” as idps. and of the Agni, 9% were idps. The Bété, also a significant ethnic group in Abidjan (9% of total sample), comprised only 6.3% of the idp sample and 5.6% of Bété respondents were idps. Table 3.3. Ethnicity of IDPs % of total sample (n=974) Baoulé

15.1%

26.0%

% of Ethnic group that are idps 17%

Guéré

5.1%

20.0%

38%

Yacouba

3.1%

13.7%

43%

Bété

9.1%

5.3%

5.6%

Agni

6.8%

6.3%

9%

% of idp sample (n=95)

Chart 3.2 shows the breakdown of the four main ethnic groups in Abidjan. The other ethnic groups comprised smaller percentages of idps, ranging from 0-5%. Chart 3.2 Ethnicity of IDPs in Abidjan 30.0%

25.0%

20.0%

% of sample % of IDPs

15.0%

10.0%

5.0%

0.0% Baoule

Guere

Yacouba

Bete

Agni

Our idp variable combined people who had been displaced from the conflict zones with those who had been displaced as a result of government forced relocation within Abidjan.When we analyzed only those who had been displaced from the conflict zones north of Abidjan, our results did not differ significantly.

32

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

4. estimates of idp population in abidjan We used our proportion of 9.7% to estimate the number of idps in Abidjan, based on the available census population estimates. The last national census was conducted in 1998, when the ten communes of Abidjan were found to have 577,258 households and 2,877,948 individuals. The official estimate for 2003 was 3,660,682 (which included the three sub prefectures outside of Abidjan City, not included in the 1998 count), and the 2006 estimate was 3,796,677.10 Using the 2006 household estimate, we divide the expected population of Abidjan by our calculated proportion of idps, to get (3,796,677/0.097) = 368,277. Using a confidence interval of ±1.85%, and our expected range of 7.9 and 11.6, we calculate that idps are in the range of 299,937–440,414. This compares with the unfpa/ensea study, conducted in 2005, which estimated a total number of 495,783 idps for Abidjan. Our estimates occurred two years later, and in the context of some return movements that had already started. Based on using a household number of 8, as indicated in our sample, the idp household estimate is thus 46,034, in the range of 37,492– 55,051. Our estimate is likely to be conservative for two reasons. First, we used a household number of 8 based on our sample findings; if we used the census number of 5, the number of households would be much higher. Second, idps are also hosted in non-idp households, but we did not include them in our estimate because unfortunately, we realized the importance of this factor only after we had begun the survey, and thus could only ask the question of a subset of respondents. One third of our sample (n=332), in four communes were asked whether any member of their household were idps and if so, how many. Of the 37 who said they were hosting idps, 11 were themselves idps according to our criteria. This means 26 or 7.8% of the sub-sample of 332 were non-idps hosting idps, as shown in Table 4.1.These non-idps hosted a total of 105 idps, or a mean number of 3 idps per non-idp household. This would add a substantial number to our estimate of idps, if not idp households.

10 Today, estimates of Abidjan’s population range from 4,225,000 (http://www.citypopulation.de/World.html) to 5 million.

33

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Table 4.1. Non-IDPs hosting IDPs by commune AtteCoMar- KouAdjame coube cody cory massi

Number of idps (%idp) Nonidps housing idps

Abobo

Yopougon

TreichVille

PortBouët

6 (7.5%)

12 (9%)

3 (4%)

7 (6%)

10 (13%)

24 (12.5%)

30 (16%)

0

3 (4%)

Not asked

1

7 (9%)

14 (12%)

Not asked

Not asked

Not asked

7 (17.5%)

8 (10%)

If we added these idps to our total idps in the sample, we would get an increased proportion of individuals in the sample, but since we are using idp households as our unit of analysis to obtain the estimate, we will simply bear this additional number in mind.

5. comparison of idps with non-idps in abidjan In this section we compare idps and non-idps in our sample, For all indicators we ran tests to determine if the two groups were statistically different. 5.1 demographics As shown in Table 5.1, idps did not differ statistically from non-idps in age or marital status. For both groups the average age of men was 37, and for women there was slight variation—35 years for non-idp women and 36 for idp women. Approximately half of the sample (49% of non-idps, 53% for idps) was married or co-habiting with another person. idp households tended to have more children. Although the table indicates that idps have slightly larger households, with an average of nine members compared with eight for non-idps, it is not a statistically significant difference.

34

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Table 5.1 Demographic comparison of IDPs and non-IDPs Non-idps (n=879) Percent male respondents 54%

idps (n=95) 54%

Mean age (Std dev: 13)

35.3

35.8

49%

53%

2.83 2.6 0-17 8.14 5.3 1-45

3.23 2.9 0-16 8.39 4.9 1-25

Marital status Number of Children

Number in household

% married or living with someone Mean Std dev Range Mean SD range

5.2 education In exploring education levels, we first split our samples by gender. We found that both idp men and women tended to have slightly more education than non-idps. As shown in Table 5.2, illiteracy rates were higher for non-idp men and women. idp men had more university education than non-idp men (44% vs. 35%). Secondary school levels were about the same, around 35%, for all groups. Larger proportions of women in both groups had only primary school. Women in both groups were significantly less likely to have had university education: only 19% of non-idp women and 12% of idp women had any university, compared with 35% (non-idp) and 44% (idp) of men. Table 5.2 Education of IDPs and non-IDPs Non-idp

idp

Women (n=398) 17%

Total (n=872) 12%

Men (n=52) 4%

Women (n=43) 14%

Total (n=95)

Illiterate

Men (n=474) 8%

Koranic/Church

4%

3%

4%

6%

0%

3%

Basic School

13%

25%

19%

8%

37%

21%

Secondary school

36%

33%

35%

37%

35%

36%

University

35%

19%

28%

44%

12%

29%

Vocatl train

1%

1%

1%

2%

0%

1%

Other

2%

2%

2%

0%

2%

1%

100%

100%

100%

100%

100%

100%

35

8%

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

5.3 housing A significant difference between idps and non-idps was their housing situation. idps were more likely to be renters than owners, and more likely to be hosted by others (usually family). Perhaps because they were living in the houses of non-idps as either renters or guests, idps and non-idps did not differ statistically in their types of housing. Most respondents (~68%) lived in housing made of bricks, and very few lived in mud houses. A slightly larger proportion of idps (9.5%) lived in temporary shelters than did non-idps (6.7%). 5.4 employment As shown in Table 5.4 and Chart 5.4, idps were more likely to be both unemployed and students, while non-idps were more likely to be “housewives” (i.e., those who worked in their homes and looked after children, but the two groups were similar in other employment categories About 17% reported full-time employment, and 17% said they were self-employed. Table 5.4 Employment in Abidjan: IDPs vs. non-IDPs Unemplyd Part-time Full-time Casl/tmp emplymnt Non-idps (n=874)

15.30%

8.00%

16.80%

2.70%

Selfemployed 18%

Students

Housewife only4

14.50%

8.40%

idps

23%

5.30%

16%

5.30%

16%

19.10%

2.10%

16.10%

7.70%

16.70%

3.00%

17.70%

15.00%

7.70%

(n=94) % total sample (n=968)

Chart 5.4 Employment in Abidjan: IDPs vs. non-IDPs Employment for IDPs and Non-IDPs 25.00%

20.00%

15.00% Non-IDPs IDPs 10.00%

5.00%

0.00% Unemplyd

Part-time

Full-time

Casl/tmp emplymnt

Selfemployed

36

Students3

Housewife only4

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

However, a notable difference was the gender similarity in the patterns of employment. As shown in Chart 5.4a, both idp and non-idp women were more likely to be self-employed than men, and much less likely to have full-time or part-time/casual employment or to be students. idp women had slightly lower rates of unemployment than either men or non-idp women. Given the relatively low numbers in each group, we cannot claim to make statistically significant claims, but the direction of our findings suggest a pattern. Chart 5.4a IDP and non-IDP Employment by Gender IDP and non-IDP Employmnt by Gender 100% 90% 80% 70% 60% Men Women

50% 40% 30% 20% 10% 0% NonIDPs

IDPs

Unempl

NonIDPs

IDPs

Full-time

NonIDPs

IDPs

P/t or casual

NonIDPs

IDPs

Self-empld

NonIDPs

IDPs

Students

5.5 living situation We examined a range of standard of living indicators, including access to potable water, distance from facilities, respondents’ expressed difficulties, and whether respondents had received assistance from the government or aid organizations in Abidjan. We found little difference in the living situation of idps compared with non-idps, although where there were differences, idps generally appeared to be worse off. The similarity of living conditions is probably explained by the fact that idps are living within host communities, sometimes even in the same houses, and while they may be slightly poorer, probably not enough to show up statistically. Access to potable water: there was no significant difference in access to potable water. Of the total sample 66% had a water connection in house, and 28.5% used water vendors. idps were slightly more likely to use water vendors and less likely to have a house connection, but this was not a statistically significant difference (see Chart 5.5).

37

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Chart 5.5 Access to potable water 80.00%

70.00%

60.00%

50.00% Non-IDPs

40.00%

IDPs

30.00%

20.00%

10.00%

0.00% House connexion

Water vendors

open well

hand pump

Illegal connexn

Access to social services was similar for both idps and non-idps. Most people (93% of the total sample) lived within a kilometer both of their children’s school and of some means of transportation. idps tended to live somewhat further away from a health facility and from a police station. Household difficulties: We asked what difficulties respondents experienced where they were living, and broke responses down by commune. We found no statistical differences between idps and non-idps. Almost all respondents mentioned the difficulties of finding work, idps were more likely to find their living area unsafe (52% versus 42%). Many respondents experienced problems with their water supply and said the drainage ditches created sanitation problems. Other widely expressed problems were distance from transportation (more than 90%). Mobility within Abidjan: idps were significantly more likely to have lived in other parts of Abidjan (57% of idps vs. 36% of non-idps).This could be a reflection of the demolition campaign and other targeting of idps groups by the government during 2002. Assistance from government or aid agency: Relatively few respondents, just 21 in total (2.2%), had ever received assistance from the government or aid agencies. idps were more likely to have received assistance, however. Of the 21 recipients, nine were idps, or 9.6% of the idp subgroup, compared with 1.4% of the non-idp subgroup.

38

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

5.6 identity documents The possession of identity documents by idps and refugees is widely viewed as a factor in protection. Lack of documents can mean people have more difficulty in establishing their right to be in a place, to claim access to services, or to lay claims to their land. In the context of “Ivoirité” in Côte d’Ivoire, birth certificates are important for accessing one’s rights as a citizen. During flight and displacement, personal documents are often lost, and idps can have difficulty replacing them. We asked our respondents to name all the identity documents they held. Most held some combination of birth certificates (93%), nationality certificates (certificat de nationalité) (72%), and national identity cards (carte d’identité nationale) (66%). Voter cards and passports were held by 18-21% of respondents. Less than 4% of the entire sample held a driver’s license. Table 5.6 shows the differences between idps and non-idps. idps were less likely to have a birth certificate (84% vs. 94% non-idps), and less likely to have consular cards. Both groups had nationality certificates, driver’s licenses and passports at similar rates. idps held national identity cards (carte d’identité nationale) and voter cards at higher rates than non-idps. Table 5.6 Possession of identity documents (percentages of those who held identity documents) Identity Document Non-idp idp Birth certificate (acte de naissance) 94% 86% Jugement supplétif 9% 8% Carte consulaire 7% 1% a Resident permit (Carte de séjour) 4% 0 Certificat de nationalité 72% 66% Carte d’identité nationale 66% 73% Voter card 21% 23% Carte de résidenceb 3% 0 Driver license 20% 18% Other (usually Passport)c 18% 18% a Carte de séjour is for residence of three months b Carte de residence is a long-term permit (10 years) for foreigners c Includes acte de notoriété valant acte de naissance

5.7 future migration: intentions to remain in or leave abidjan We asked respondents whether they and their families hoped to return to their place of habitual residence, resettle elsewhere in the country or remain in Abidjan. As shown in Table 5.7 and Chart 5.7, there were

39

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

several significant differences between idps and non-idps. idps were more likely to want to return home than non-idps (44% vs. 4%), and less likely to want to remain where they were in Abidjan. Whereas 77% of non-idps wanted either to remain where they were or move elsewhere in Abidjan, this was the case for only 34% of idps. Of all our respondents only two (less than 1%) said they wanted to migrate to North America or Europe. Table 5.7 Intentions about future mobility/migration Remain in Settle Return Settle current area elsewhere in home elsewhere in Abidjan Abidjan Côte d’Ivoire Nonidp

idp

48% 16%

30% 22%

4% 44%

3% 2%

Go to another country in Africa 4% 1%

Pearson Chi = 226.793, signif 0.00 Chart 5.7 Intentions about future mobility/migration 60.00%

50.00%

40.00%

% NonIDP %IDPs

30.00%

20.00%

10.00%

0.00% Return home

Resettle elsewhere in Abidjan

Resettle elsewhere in Côte d’Ivoire

Go to other country in Africa

Don’t Remain in current area know/Not yet decided

5.8 validity check Our enumerators had extensive experience working with idps in Abidjan, and in focus groups we asked them to comment on our findings. Their view was that our data reflect the situation of idps in Abidjan, and that the problems mentioned by idps, such as adapting to urban life, finding transportation and accessing social networks that could help with finding a job, were not different from the difficulties any newcomer to the city would face. Our methodology did not lend itself to probing these problems more deeply, nor did or survey explore whether idps experienced greater language barriers, or whether idps

40

Go to US or Eur <1% 0

Don’t Know

11% 15%

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

were subject to higher prices for housing, food, and living expenses. In general, idps appeared to find urban life very expensive, and needed at least two years to adapt. Notable in our findings was the absence of reference to security problems, beyond those related to crime.

6. analysis and conclusion For the most part the people we identified as internally displaced appear to be relatively well integrated in Abidjan. One explanation is that many of the idps who came to Abidjan were sympathetic to the government authorities, while those who sympathized with the rebels remained in the conflict zone. We did not ask about political sympathies in our survey, or whether people felt they had been persecuted for political or ethnic reasons. But similarities with the Abidjan urban population and possible sympathies with the government might explain why the idps in Abidjan did not experience much hostility or discrimination. It might also indicate that ethnic or political cleansing occurred in the conflict areas, but we have no data on this. However while idps and non-idps experience the same stresses related to urban poverty and lack of adequate infrastructure, there were notable differences in their employment and housing situations, in the kinds of identity documents they possess, and in their desire to leave Abidjan. These differences suggest that idps are more vulnerable; they experience more structural insecurity and find it harder to lay claim to their rights as citizens. Key areas of vulnerability include identity documents, housing and employment, where idps appear to have more difficulties than non-idps. It is worth exploring further to see whether idps are being discriminated against, when it comes to housing and employment. The problem of obtaining or replacing birth certificates and other ID is clearly an area of programming that aid agencies could work with the government to address (some have begun to do this in cooperation with the government). Our study constructed a category called “idps” that comprised people who shared certain commonalities and who all had said they had been displaced. It is likely that we did not capture everyone in the sample who would qualify as an idp according to the Guiding Principles on Internal Displacement. 11Aside from the possibility that respondents may have been reluctant to acknowledge certain aspects of their situation, our measures might have missed some idps because of the complexity of their migration movements. 11 http://www.brookings.edu/projects/idp/gp_page.aspx

41

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

Our study only partially captures the complexity of population movements during and after a conflict, when the reasons for coming to a particular place are often different from those that led to the initial displacement. When armed conflict erupts in the areas where economic conditions are deteriorating, people are likely to migrate to urban centers after they are initially displaced. This subsequent urban migration may occur at a later date, and it is prompted by different reasons than those that caused the initial displacement. Thus a person may come to Abidjan to find work, but might have initially been forcibly displaced as a result of being forced off family land, or fleeing violence. The conflict in Ivory Coast erupted in a context of liberalization of the economy, socio-economic marginalization of hundreds of thousands of people, and politicization of ethnicities. These factors triggered or unleashed migration to Abidjan that would have taken place anyway, but at a slower pace and along the lines of global urbanization trends. Forcibly displaced people, or idps, were clearly among those who would have migrated to Abidjan anyway. But urban idps are worse off before their migration, during their journeys and, as our study suggests, after their arrival. Their plight warrants attention by humanitarian agencies. We believe that it is possible to capture the complexity of the idp situation conceptually, and at the same time provide governments and other institutions with analytical tools and data that help them find ways to protect the rights of idps. Our study sought to develop an indicator for idps in the context of complex migration patterns and the absence of a set of clear-cut inclusion/exclusion criteria for identifying idps. The Guiding Principles offer guidance, but do not provide measurable indicators that can be used in an empirical study. Our approach sought to avoid both over-simplifying and over-complicating the concept of urban idps by using criteria that capture the different aspects of forced displacement. In other contexts, these criteria can be adapted. There are no quick-fix solutions to the achievement of durable solutions for the idps and/or conflict-affected people in the areas we surveyed. However our framework and findings can be used as a starting point from which to analyze and compare socio-economic conditions and prospects in Abidjan with those in areas of origin. This comparison, combined with a political analysis at national and local levels/areas of origin, could form the basis for programming of durable solutions, whether those solutions are supporting return, local integration or resettlement.

42

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

appendix a: abidjan communes and quartiers with no. of survey respondents Commune

Quartier

Sample size (no. of households)

Jean P II Nemantoulay Locodjoi Sante III

20 20 20 20

Cocody N=121

Angre Lycée technique Université Aghien Akuedo Vill Danga

20 20 40 20 1 20

Marcory N=79

Aliodan Anoumabo Gnanzoua Maire Koré

20 19 20 20

Koumassi N=121

EMCC Angre Prodomo SpoPan Nord-Est Mosqué Grand Marché Sicogi III

20 20 21 20 20 2 1

Adjamé N=79

Marie Williamsville Bromoké ? ou Pelieuville? Adjamé-Nord

20 20 20 19

Abobo N=192

Abobo-Nord Set? Abobo-Te Agbekoi Agnissankoi Avocat Djibi Segbe Ctr AnonkoiIII Houphoët Boigny Sagbe Sud Akeikoi

20 20 20 19 13 20 20 20 20 20

Yopougon N=183

Gare-Sud Sodeci Koute Vill Port Boiute II Sideko Sicogi Loc Sogefina-Koute BancoII Gesco Manuten Maine Niangon Sud Zone Indust

19 1 20 20 19 20 20 20 24 20

Treichville

Kouassi-Len Cite Function Anani Anam Hopital Grand march Vridi Gendarmerie Vridi-Canal Sir Adjoufou I

20 20 1 20 20 19 19

Attecoube N=80

Port-Bouet N=79

43

Internal Displacement to Urban Areas: the Tufts-IDMC Profiling Study. Abidjan, Côte d’Ivoire: Case 2 Jacobsen, with the Internal Displacement Monitoring Centre

appendix b: ethnicity of idps and non-idps in abidjan Respndnt Ethnic group Baoulé

Non-

IDP

Total

%IDP

22

148

15%

Mossi

18

18

0%

100%

Malinke

44

45

2%

98%

Peulh

14

14

0%

100%

Senoufo

43

Guéré

31

4

47

9%

91%

19

50

38%

62%

Bété

84

5

89

6%

94%

Agni

62

5

67

7%

93%

Yacouba

21

9

30

30%

70%

More

10

10

0%

100%

Abron

17

17

0%

100%

Gouro

36

38

5%

95%

IDP

126

1

2

% NonIDP

85%

Adjoukrou

14

2

16

13%

88%

Dioula

54

1

55

2%

98%

Dida

25

1

26

4%

96%

Attie

51

2

53

4%

96%

Appolo

17

17

0%

100%

Aboure

12

12

0%

100%

Ebrie

31

31

0%

100%

Abbay

21

23

9%

91%

8%

92%

2

Other

69

6

75

Total

800

81

881

44